Abstract Model updating of dynamical systems has been attracting much attention because it a very wide range applications in aerospace, civil, and mechanical engineering, etc. Many methods were developed there substantial development Bayesian for this purpose the recent decade. This article introduces some state-of-the-art work. It consists two main streams model updating, namely using response time history modal measurements. The former one utilizes directly histories identification...

Summary In this paper, a Bayesian sequential sensor placement algorithm, based on the robust information entropy, is proposed for multi‐type of sensors. The presented methodology has two salient features. It holistic approach such that overall performance various types sensors at different locations assessed. Therefore, it provides rational and effective strategy to design configuration, which optimizes use available resources. This algorithm very efficient due its nature, in prior...

This paper reports the structural health monitoring benchmark study results for Canton Tower using Bayesian methods. In this study, output-only modal identification and finite element model updating are considered a given set of acceleration measurements corresponding ambient conditions 24 hours. first stage, spectral density approach is used with time histories as excitation to tower unknown. The parameters associated uncertainty can be estimated through inference. Uncertainty...

A Bayesian probabilistic method is proposed for online estimation of the process noise and measurement parameters Kalman filter. filter a well-known recursive algorithm state dynamical systems. In this algorithm, it required to prescribe covariance matrices noise. However, inappropriate choice these substantially deteriorates performance paper, which govern matrices. The not only estimates optimal but also quantifies associated uncertainty in an manner. By utilizing estimated parameters,...

Abstract In this article, a novel Bayesian framework is proposed for real‐time system identification with calibratable model classes. This self‐calibrating scheme adaptively reconfigures the classes to achieve reliable estimation state and parameters. At each time step, plausibilities of are computed they serve as cue calibration. Once calibration triggered, all will be reconfigured. Thereafter, continue propagate calibrated until next recalibration. Consequently, evolve their deficiencies...

Seismic response prediction is a challenging problem and significant in every stage during structure's life cycle. Deep neural network has proven to be an efficient tool the of structures. However, conventional with deterministic parameters unable predict random dynamic In this paper, deep Bayesian convolutional proposed seismic response. The Bayes-backpropagation algorithm applied train learning model. A numerical example three-dimensional building structure utilized validate performance...

Elasto-magnetic (EM) methods have been widely adopted in civil engineering for the tension monitoring of cables, hangers, and prestressed strands due to their noncontact nondestructive properties. However, practical applications, variations installation positions EM sensor often lead significant differences wire length. This results changes resistance, which can affect measurement results. Therefore, this study, we developed a multiphysics coupling finite element model systematically analyze...

In this paper, a stable and robust filter is proposed for structural identification. This resolves the instability problems of traditional extended Kalman (EKF). Instead ad hoc assignment noise covariance matrices in EKF, (SREKF) provides real-time updating parameters. well-known problem EKF due to improper matrices. Furthermore, SREKF capable removing abnormal data points manner. As result, parametric identification results will be more reliable have fewer fluctuations. The approach applied...

Abstract Spatial modeling is a core element in geographical information science. It incorporates geographic to construct the relationship for interpreting behavior of spatial phenomena. In this paper, broad learning framework nonparametric presented. Broad overcomes obstacle expensive computational consumption deep and provides powerful computationally efficient alternative. contrast architecture that configured with stacks hierarchical layers, networks are established flat manner can be...

Automatic crack identification for pipeline analysis utilizes three-dimensional (3D) image technology to improve the accuracy and reliability of identification. A new technique that integrates a deep learning algorithm 3D shadow modeling (3D-SM) is proposed automatic corrosion cracks in pipelines. Since depth below surrounding area crack, projected when exposed under light sources. In this study, we analyze areas through identify evolving shape shadows. To denoise images, connected domain...

In this study, the Bayesian probabilistic framework is investigated for modal identification and identifiability based on field measurements provided in structural health monitoring benchmark problem of an instrumented cable-stayed bridge named Ting Kau Bridge (TKB). The comprehensive system TKB has been operated more than ten years it recognized as one best test-beds with readily available measurements. established to stimulate investigations present paper addresses from prospective....

There has been a vision of creating bridge digital twins as virtual simulation models assets to facilitate remote management. Bridge model updating is one twin technology which can enable the continuous structural new monitoring data collected. This paper examines why there currently little industry uptake monitoring, modelling and for operation maintenance bridges despite over two decades research in these fields. The study analyses findings from series semi-structured interviews with...

Damage detection of civil and mechanical structures based on measured modal parameters using model updating schemes has received increasing attention in recent years. In this study, for uncertainty-oriented damage identification, a non-probabilistic structural identification (NSDI) technique an optimization algorithm interval mathematics is proposed. order to take into account the uncertainty quantification, elastic modulus described as unknown-but-bounded values proposed new scheme...

Abstract In this paper, a novel telescopic broad Bayesian learning (TBBL) is proposed for sequential learning. Conventional suffers from the singularity problem induced by complexity explosion as data are accumulated. The TBBL successfully overcomes challenging issue and feasible with big streams. network of reconfigurable to adopt augmentation condensation. As time evolves, augmented incorporate newly available additional components. Meanwhile, condensed eliminate connections components...